1. Purpose: In this study, we propose to 1) identify the night-to-night variability of the plasma dim light melatonin onset (DLMO) in elderly and 2) determine the role of salivary melatonin and urine melatonin metabolites as surrogate markers for the plasma. This information is crucial to the development of the DLMO as a fundamental tool for investigation of circadian phase abnormalities. We will select a group of 15 elderly subjects (age>65) with primary insomnia and 15 elderly subjects without insomnia or other sleep complaints. 2. Background: Primary insomnia is a common clinical complaint in the elderly 45%-15%) and has a significant impact on their quality of life as well as daytime alertness. While the etiology of primary insomnia has remained elusive, preliminary research has suggested that abnormalities of the deep circadian play a role: phase advance of the circadian pacemaker leading to a relative desynchrony between the intrinsic desire to sleep and external cues, and impaired amplitude of the circadian pacemaker with inadequate drive to sleep. One tool to measure circadian rhythms is the onset of melatonin under dim light conditions (bright light blunts melatonin secretion)-the dim light melatonin onset (DLMO). However, the applicability of the DLMO in the elderly insomniac population is a topic which remains unclear for the following reasons: 1) Virtually all DLMO studies to date have been conducted in young, healthy subjects, 2) These studies have not investigated both cases with insomnia and controls, and 3) There is little data concerning the night-to-night variability of the plasma DLMO, an issue which is of paramount importance if the DLMO is to be considered a stable marker of the circadian phase.